Search

With the rapid advancement of 5G mobile networks and Internet of Things technology, an increasing number of data-intensive applications are generating massive amounts of information, such as face recognition, video stream analysis, and augmented reality. These applications not only demand significant computational resources but also require high real-time performance, posing challenges to the existing cloud-computing model. Deploying these data-intensive applications in a mobile edge computing environment can reduce response time for processing user tasks and meet low-latency requirements. However, conventional CPU scheduling strategies fail to effectively enhance the performance of data-intensive applications on edge nodes with limited computing resources. In this study, we focus on real-time application feature awareness and propose a strategy considering data arrival rate and cost fair CPU scheduling (RCFS) with two components: CPU resource allocation based on process weights and process scheduling based on distributed weighted round-robin. In compared to the default scheduling strategy in Linux, named Completely Fair Scheduler, experiments on edge nodes show that the RCFS strategy can effectively enhance CPU utilization, reduce running time of data-intensive applications, and improve the system's data throughput. In the best-case scenario, the RCFS strategy achieved a remarkable increase in CPU utilization of data-intensive applications by up to 98.5%, reduced running time by up to 45.5%, and improved data throughput by up to 83.6%. [ABSTRACT FROM AUTHOR]

Heterogeneous computer systems are becoming mainstream due to their disparate processing and performance capabilities. These systems consist of different types of devices, i.e., central processing units (CPUs), accelerators, and graphics processing units (GPUs). In the heterogeneous computing environment, if one device is more powerful in terms of computing capability, the scheduling schemes generally favor the powerful device, and that device becomes overloaded, while the other device is underutilized. This load imbalance problem results in increased execution time. In this research, we propose load-balanced task scheduler combined with machine learning-based device predictor. The device predictor is used to predict execution time both on CPU and GPU devices, and a device with shorter predicted execution time is considered as a suitable device for that particular task. However, it may happen that a high fraction of tasks map only on one type of device since that device is considered as a suitable device for them. It is due to the fact that a task is mapped to one device (with lower predicted execution time), although it can be executed on the other device as well. In this context, one device may become overloaded, while the other device may be underutilized. To solve this problem of load imbalance, we use work-stealing-based task scheduler as part of our solution that allows an idle device to process tasks from the queue of another's device. In this way, we can avoid load imbalance, minimize the overall execution time of tasks, and maximize the device utilization and throughput. We evaluate the performance of our proposed solution into two stages. Firstly, we measure the error rate of our machine learning predictor using three different algorithms (i.e., random forest, gradient boosting, and multiple linear regression). We demonstrate that random forest performs better with marginal error rate. Secondly, we compare the performance of work-stealing task scheduler with other scheduling alternatives. Our results show that the proposed solution reduces execution time by 65.63%, increased resource utilization by 93.3%, and throughput by 65.5% in comparison with baseline scheduling schemes. [ABSTRACT FROM AUTHOR]

Recently, wide-area distributed computing environments have become popular owing to their huge resource capability. In a wide-area distributed computing environment, joint scheduling of tasks and data is the main strategy to improve system performance. However, the geographically distributed diverse resources exhibit high variations, making it challenging to design efficient joint scheduling of tasks and data. To accurately adapt to the dynamic variations of geographically distributed diverse resources and achieve a high system performance, this study proposes a hypergraph-partitioning-based online joint scheduling method. The proposed method constructs a hypergraph of geographically distributed tasks, data, and diverse resources to clearly describe the correlation among the three elements and quantitatively reflect the time cost of different process in the environment. The hypergraph is dynamically updated according to the generated scheduling scheme and the collected information to reflect the dynamic variations of resource states. Then, a hypergraph partition optimization mechanism is proposed to generate efficient joint scheduling schemes, thus reducing the overall completion time in the system. The experimental results indicate that compared with the state-of-the-art joint scheduling methods, the proposed method reduces the overall completion time by up to 25.67% and significantly reduces the task waiting time, although it makes a concession in the data migration time. [ABSTRACT FROM AUTHOR]

The off-line scheduling scheme of existing wireless sensor network is improved in this paper. Firstly, Bayesian method is introduced and Poisson distribution parameter followed by the number of sensor nodes is regarded as a random variable. We estimate the parameter by Bayesian posteriori estimate and obtain Bayesian estimated correction value. Then, we discuss the relationship between instantaneous event capture probability, average capture probability, average capture energy efficiency and Bayesian estimate value of the distribution parameter. Finally, considering the fact that Bayesian estimation value can be adjusted automatically after posterior samples are included, we propose an online scheduling scheme for asynchronous wireless sensor network. Through simulation comparison between online scheduling scheme and off-line scheduling scheme, our results show that online scheduling scheme can lower the probability of failure to capture stochastic events, increase the probability of capturing events and further save the energy of wireless sensor network, and be more flexible to capturing stochastic events. Moreover, expanding the perception radius of sensors can also enhance capture efficiency on the basis of controlling the working duration of wireless sensor network. [ABSTRACT FROM AUTHOR]

Heterogeneous multi-core processors (HMP) are dual-objective hardware platforms which integrate both high-performance and low power consumption processors. Investigation of simultaneous fairness and energy efficiency could widen their applicability and reveal their strengths. This paper proposes a scheduling framework considering energy efficiency, shared resource contention, and fairness for heterogeneous multi-core processors. The presented framework is implemented and evaluated on a real HMP platform. The obtained experimental results via SPEC CPU2006 benchmark indicate that the proposed framework surpasses Linux and four other schedulers in terms of fairness (58% on average) and energy efficiency (37% on average). The source code of the proposed framework and the opponent algorithms are available online at https://github.com/baghers/CEEF. [ABSTRACT FROM AUTHOR]

In partially run-time reconfigurable (PRR) FPGAs, hardware tasks should be configured before their execution. The configuration delay imposed by the reconfiguration process increases the total execution time of the hardware tasks and task graphs. In this paper, a new technique named forefront-fetch is presented to improve the makespan of hardware task graphs running on PRR FPGAs via alleviating the adverse effects of the configuration delays. In this technique, which is applied to a sequence of task graphs, the configuration of some tasks is carried out within the execution phase of the previous task graph. This strategy leads to hide the configuration delay of the forefront-fetched tasks that as a result improves the execution time. The proposed solution modifies the schedules of the task graphs at design time to obtain a set of schedule pairs for the run-time environment. Experiments on actual and synthesized task graphs demonstrate the ability of the proposed technique in improving the makespan of hardware task graphs. The obtained results show that for a set of task graphs running on Xilinx™ Virtex-5 XUPV5LX110T FPGA, makespan is improved by 37.81% on average. Moreover, the proposed solution outperforms state-of-the-art prefetch-aware scheduling strategies by 14.78% makespan improvement. [ABSTRACT FROM AUTHOR]

MapReduce framework is an effective method for big data parallel processing. Enhancing the performance of MapReduce clusters, along with reducing their job execution time, is a fundamental challenge to this approach. In fact, one is faced with two challenges here: how to maximize the execution overlap between jobs and how to create an optimum job scheduling. Accordingly, one of the most critical challenges to achieving these goals is developing a precise model to estimate the job execution time due to the large number and high volume of the submitted jobs, limited consumable resources, and the need for proper Hadoop configuration. This paper presents a model based on MapReduce phases for predicting the execution time of jobs in a heterogeneous cluster. Moreover, a novel heuristic method is designed, which significantly reduces the makespan of the jobs. In this method, first by providing the job profiling tool, we obtain the execution details of the MapReduce phases through log analysis. Then, using machine learning methods and statistical analysis, we propose a relevant model to predict runtime. Finally, another tool called job submission and monitoring tool is used for calculating makespan. Different experiments were conducted on the benchmarks under identical conditions for all jobs. The results show that the average makespan speedup for the proposed method was higher than an unoptimized case. [ABSTRACT FROM AUTHOR]

As a new type of computing, cloud computing has led to a major computational change. Among many technologies in cloud computing, task scheduling has always been studied as a core issue by industry and academia. In the existing research, the main goal is completion time or load balancing. However, as the expansion of cluster size, energy consumption becomes a problem that must be faced. In this paper, the first of maximum loss scheduling algorithm is proposed. The algorithm is a low-power algorithm that can greatly reduce the energy consumption of cloud computing clusters through loss comparison rule. The effect of this method is more obvious as the cluster size and the number of tasks increase. Experimental simulation results show that the proposed method is significantly better than the Max–Min, Min–Min, Sufferage and E-HEFT algorithms. Compared to Min–Min, Max–Min, Sufferage and E-HEFT algorithms, average completion time of the algorithm reduces 16%, 12%, 8% and 14%, respectively. At the same time, the load balancing effect is also better than Min–Min and Sufferage algorithms. [ABSTRACT FROM AUTHOR]

The development of Internet of Things leads to an increase in edge devices, and the traditional cloud is unable to meet the demands of the low latency of numerous devices in edge area. On the hand, the media delivery requires high-quality solution to meet ever-increasing user demands. The edge cloud paradigm is put forward to address the issues, which facilitates edge devices to acquire resources dynamically and rapidly from nearby places. However, in order to complete as many tasks as possible in a limited time to meet the needs of users, and to complete the consistency maintenance in as short a time as possible, a two-level scheduling optimization scheme in an edge cloud environment is proposed. The first-level scheduling is by using our proposed artificial fish swarm-based job scheduling method, most jobs will be scheduled to edge data centers. If the edge data center does not have enough resource to complete, the job will be scheduled to centralized cloud data center. Subsequently, the job is divided into same-sized tasks. Then, the second-level scheduling, considering balance load of nodes, the edge cloud task scheduling is proposed to decrease completion time, while the centralized cloud task scheduling is presented to reduce total cost. The experimental results show that our proposed scheme performs better in terms of minimizing latency and completion time, and cutting down total cost. [ABSTRACT FROM AUTHOR]

Reconfigurable devices (RDs) are extremely advantageous when employed in real-time embedded systems. Nonetheless, they are susceptible to soft errors. In a broad sense, the present research addresses the challenge of improving the reliability of independent periodic real-time hardware tasks in RDs by utilizing hybrid fault-tolerant scheduling. The current paper combines static and dynamic real-time scheduling techniques to improve the reliability of the system. First, the proposed algorithm statically schedules primary tasks and preserves area and time for possible backup tasks on the RD. The overlapping of passive backup tasks is possible. Next, at the run time, event-triggered dispatcher dynamically determines which candidate backup copy should be selected for configuration on the overloaded preserved areas. Reliability, task deadline, and RD area limitations are the determining factors of backup overloading in the static phase. On the other hand, in the dynamic phase, the execution result of the primary tasks—in this case, success or failure—is the deciding factor based on which the dispatcher configures the true backup task on the preserved area. Experimental results show that the hybrid scheduling technique enhances the mean-time-to-failure of the system by an average factor of 1.22 in comparison with a similar state-of-the-art study. [ABSTRACT FROM AUTHOR]

The ever-increasing complexity in terms of heterogeneity and available parallelism of modern architectures, together with the growing software ecosystem in terms of application libraries and runtime systems are posing a new challenge in terms of programmability on top of the already difficult task of extracting high parallel performance. With the goal of simplifying the programmability at the extreme, we first present the single task–single executor (STSE) model, in which the execution entry point consists solely on a single task assigned to an abstract execution context or executor, representing the whole system platform. As a second contribution, we introduce the expanded scheduling operations that endow executors with additional runtime task scheduling capabilities, and explain how executors can be composed to express and exploit different architectural features at run time under the framework of a single/multiple–task/executor taxonomy. As a result, we introduce STEEL, a novel programming model built on these concepts and STEEL-RT, a coupled high-performance runtime implementation for it. We illustrate the potential of STEEL by means of runtime experiments on a modern heterogeneous architecture, showing that complex and efficient heterogeneous executions are accessible from a STSE approach. [ABSTRACT FROM AUTHOR]

Graphics processing units (GPUs) have become one of the best platforms for exploiting the plentiful thread-level parallelism of applications. However, GPUs continue to underutilize their hardware resources for optimizing the performance of numerous general-purpose applications. One primary reason for this is the inefficiency of existing warp schedulers in hiding long-latency operations such as global loads and stores. This study proposes a long-latency operation-based warp scheduler to improve GPU performance. In the proposed warp scheduler, warps are partitioned into different pools based on the characteristics of instructions that are subsequently executed. Specifically, this warp scheduler uses warps that are likely waiting for long-latency operations for a guiding role. Meanwhile, other warps perform filling roles (i.e., to overlap the latencies caused by the guiding warps). Our experimental results demonstrate that the proposed warp scheduler improves GPU performance by 24.4% on average as compared to the conventional warp scheduler. [ABSTRACT FROM AUTHOR]

Multiple CPUs and GPUs are integrated on the same chip to share memory, and access requests between cores are interfering with each other. Memory requests from the GPU seriously interfere with the CPU memory access performance. Requests between multiple CPUs are intertwined when accessing memory, and its performance is greatly affected. The difference in access latency between GPU cores increases the average latency of memory accesses. In order to solve the problems encountered in the shared memory of heterogeneous multi-core systems, we propose a step-by-step memory scheduling strategy, which improve the system performance. The step-by-step memory scheduling strategy first creates a new memory request queue based on the request source and isolates the CPU requests from the GPU requests when the memory controller receives the memory request, thereby preventing the GPU request from interfering with the CPU request. Then, for the CPU request queue, a dynamic bank partitioning strategy is implemented, which dynamically maps it to different bank sets according to different memory characteristics of the application, and eliminates memory request interference of multiple CPU applications without affecting bank-level parallelism. Finally, for the GPU request queue, the criticality is introduced to measure the difference of the memory access latency between the cores. Based on the first ready-first come first served strategy, we implemented criticality-aware memory scheduling to balance the locality and criticality of application access. [ABSTRACT FROM AUTHOR]

Since scientific workflow scheduling becomes a major energy contributor in clouds, much attention has been paid to reduce the energy consumed by workflows. This paper considers a multi-objective workflow scheduling problem with the budget constraint. Most existing works of budget-constrained workflow scheduling cannot always satisfy the budget constraint and guarantee the feasibility of solutions. Instead, they discuss the success rate in the experiments. Only a few works can always figure out feasible solutions. These methods work, but they are too complicated. In workflow scheduling, it has been a trend to consider more than one objective. However, the weight selection is usually ignored in these works. The inappropriate weights will reduce the quality of solutions. In this paper, we propose an energy-aware multi-objective reinforcement learning (EnMORL) algorithm. We design a much simpler method to ensure the feasibility of solutions. This method is based on the remaining cheapest budget. Reinforcement learning based on the Chebyshev scalarization function is a new framework, which is effective in solving the weight selection problem. Therefore, we design EnMORL based on it. Our goal is to minimize the makespan and energy consumption of the workflow. Finally, we compare EnMORL with two state-of-the-art multi-objective meta-heuristics in the case of four different workflows. The results show that our proposed EnMORL outperforms these existing methods. [ABSTRACT FROM AUTHOR]

The performance of scheduling algorithms for HPC jobs highly depends on the accuracy of job runtime values. Prior research has established that neither user-provided runtimes nor system-generated runtime predictions are accurate. We propose a new scheduling platform that performs well in spite of runtime uncertainties. The key observation that we use for building our platform is the fact that two important classes of scheduling strategies (backfilling and plan based) differ in terms of sensitivity to runtime accuracy. We first confirm this observation by performing trace-based simulations to characterize the sensitivity of different scheduling strategies to job runtime accuracy. We then apply gradient boosting tree regression as a meta-learning approach to estimate the reliability of the system-generated job runtimes. The estimated prediction reliability of job runtimes is then used to choose a specific class of scheduling algorithm. Our hybrid scheduling platform uses a plan-based scheduling strategy for jobs with high expected runtime accuracy and backfills the remaining jobs on top of the planned jobs. While resource sharing is used to minimize fragmentation of resources, a specific ratio of CPU cores is reserved for backfilling of less predictable jobs to avoid starvation of these jobs. This ratio is adapted dynamically based on the resource requirement ratio of predictable jobs among recently submitted jobs. We perform extensive trace-driven simulations on real-world production traces to show that our hybrid scheduling platform outperforms both pure backfilling and pure plan-based scheduling algorithms. [ABSTRACT FROM AUTHOR]

Large-scale computing platforms become essential in nowadays business and scientific activities. The electrical energy consumed by such platforms increases dramatically due to the increase in both the computation power of these platforms and the required cooling energy. Task scheduling is one of the key issues to achieve high performance in large-scale computing platforms. One approach to the compile-time task scheduling, in these platforms, is duplication-based list scheduling heuristics. This approach enhances the performance from the application completion-time point of view, while it decreases the performance from the energy consumption point of view. This paper presents a mechanism that can be applied to any schedule produced by a duplication-based task scheduling algorithm to reduce the consumed energy while keeping the same application completion time. The mechanism is called toward green task duplication (TGTD). TGTD is examined in reducing the energy consumed by four of the most-recent and well-known duplication-based list scheduling algorithms. The experimental results based on a computer simulation utilizing large sets of both randomly generated and two real-world applications graphs show that TGTD significantly enhances the energy consumed by each algorithm. [ABSTRACT FROM AUTHOR]

The paper studies the performance of algorithms for scheduling of many-task applications in distributed computing systems. Two important classes of such applications are considered: bags-of-tasks and workflows. The comparison of algorithms is performed on the basis of discrete-event simulation for various application cases and system configurations. The developed simulation framework based on SimGrid toolkit provides the necessary tools for implementation of scheduling algorithms, generation of synthetic systems and applications, execution of simulation experiments and analysis of results. This allowed to perform a large number of experiments in a reasonable amount of time and to ensure reproducible results. The conducted experiments demonstrate the dependence of the performance of studied algorithms on various application and system characteristics. While confirming the performance advantage of advanced static algorithms, the presented results reveal some interesting insights. In particular, the accuracy of the used network model helped to demonstrate the limitations of simple analytical models for scheduling of data-intensive parallel applications with static algorithms. [ABSTRACT FROM AUTHOR]

"More data, more information." Big data helps businesses and research communities to gain insights and increase productivity. Many public cloud service providers offer Hadoop MapReduce as a service based on pay-per-use via infrastructure as a service on clusters of virtual machines promising on-demand horizontal scaling. These clusters of virtual machines are launched in various physical machines across racks in cloud data centers. Such multi-tenancy negatively introduces performance heterogeneity for Hadoop virtual machines due to hardware heterogeneity and interference from co-located virtual machine. Performance heterogeneity largely affects MapReduce job latency and resource utilization of rented Hadoop virtual clusters. Default MapReduce schedulers assign map/reduce tasks assuming the hardware is homogeneous. Interference-aware schedulers perform by only observing the interference pattern generated by co-located virtual machines. These schedulers do not consider the heterogeneous performance of virtual machines. Therefore, we propose a dynamic ranking-based MapReduce job scheduler that places the map and reduces tasks based on a virtual machine's performance rank to minimize job latency and improve resource utilization. Our proposed approach calculates the performance score for each virtual machine based on hardware heterogeneity and co-located virtual machine interference. Then, it ranks the virtual machines based on the map and reduce performance separately to place map and reduce tasks. To demonstrate our ideas, we have set a test bed with 29 virtual machines on eight physical machines with different configurations and capacities. We modify a default fair scheduler in Hadoop 2.x to incorporate our ideas and evaluate them with different workloads on the PUMA dataset. The proposed method is then compared against a default fair scheduler (resource-aware) and an interference-aware scheduler based on job latency and resource utilization. Finally, we argue in favor of our approach as it improves resource utilization by 30–65% and overall job latency by up to 30%. [ABSTRACT FROM AUTHOR]

In the current decade, doing the search on massive data to find "hidden" and valuable information within it is growing. This search can result in heavy processing on considerable data, leading to the development of solutions to process such huge information based on distributed and parallel processing. Among all the parallel programming models, one that gains a lot of popularity is MapReduce. The goal of this paper is to survey researches conducted on the MapReduce framework in the context of its open-source implementation, Hadoop, in order to summarize and report the wide topic area at the infrastructure level. We managed to do a systematic review based on the prevalent topics dealing with MapReduce in seven areas: (1) performance; (2) job/task scheduling; (3) load balancing; (4) resource provisioning; (5) fault tolerance in terms of availability and reliability; (6) security; and (7) energy efficiency. We run our study by doing a quantitative and qualitative evaluation of the research publications' trend which is published between January 1, 2014, and November 1, 2017. Since the MapReduce is a challenge-prone area for researchers who fall off to work and extend with, this work is a useful guideline for getting feedback and starting research. [ABSTRACT FROM AUTHOR]

Recently, Service-level agreement (SLA) is deemed to be an integral aspect for on-demand provisioning of scalable resources on Cloud. SLA defines important constraints for instance guaranteed quality of service (QoS), pricing, fault-tolerant availability, security, and period of service. Currently, there is a dire need of SLA-based scheduling that improves resources utilization on Cloud. Scheduling with reduced execution time and cost may adversely affect resource utilization. To overcome this issue, we present a cost-efficient SLA-based load balancing scheduler, namely SLA-RALBA, for heterogeneous Cloud infrastructures. The proposed technique supports three levels of SLA opted by the Cloud users. The proposed novel technique incorporates the execution cost for the successful execution of users' services to elevate the resource utilization on Cloud. The SLA-RALBA is simulated for performance analysis using the benchmark GoCJ and HCSP datasets. The performance results of the SLA-RALBA are compared with the existing schedulers, namely Execution-MCT, Profit-MCT, SLA-MCT, Execution-Min-Min, Profit-Min-Min, and SLA-Min-Min in terms of average resource utilization, execution time, and cost of the Cloud services. The obtained results reveal that SLA-RALBA provides an even trade-off between execution time and cost of the services by guaranteeing a drastic improvement in resource utilization on Cloud than existing algorithms. [ABSTRACT FROM AUTHOR]

Scheduling in cloud computing is the assignment of tasks to resources with maximum performance, which is a multi-purpose problem. The scheduling is of NP-Hard issues that is the reason why meta-heuristic algorithms are used in scheduling problems. The meta-heuristic scheduling algorithms are divided into two categories of biological and non-biological. Swarm-based meta-heuristics are of biological algorithms that are based on imitation, or based on sign. The whale optimization algorithm is a meta-heuristic biological swarm-based intelligence algorithm based on imitation. This algorithm suffers from the early convergence problem which means the population convergences early to an unfavorable optimum point. Usually, the early convergence occurs because of the weakness in exploration capability (global search). In this study, an optimized version of the Whale optimization algorithm is introduced that presents a new idea in grouping whales called GWOA. It is firstly proposed to overcome the early convergence problem and then make a balance between the local and the global search in finding the optimal solution. The proposed method divides the sorted population into δ groups and a member of each group is randomly selected which is used in encircling prey section of the whale optimization algorithm. Then, the average best fitness was enhanced to improve both exploitation and exploration as well as premature convergence. In the next step, GWOA is used in a cloud computing scheduler at high workload to reduce the average execution time, response time, and increase the throughput in the cloud computing environment. The proposed whale optimization algorithm is compared with the standard whale optimization algorithm (WOA), improved whale optimization algorithm (CWOA), particle swarm optimization (PSO), and bat algorithms applying CEC2017 functions to compare the average parameter of the best fit, and then they are implemented as a cloud computing scheduler. The results of the experiments show that the proposed method has a better performance in comparison with competent meta-heuristic algorithms and scheduling algorithms. [ABSTRACT FROM AUTHOR]

Energy consumption and elevating the chip temperature become a serious challenge in designing embedded systems mainly due to transistor scaling and integration of more components into a single chip. The temperature of the chip has significant effects on leakage current, energy consumption and reliability of the chip; hence, discovering a mechanism that reduces both energy consumption and temperature of the chip is of utmost significance. An effective task scheduling in a real-time multiprocessor system-on-chip system has a direct impact on energy consumption and temperature of the chip. Several task scheduling and task assignment techniques have been proposed to achieve this goal. Most of those works consider only utilization of processors to distribute tasks among processors in order to reduce energy consumption. Meanwhile, those works try to reduce temperature in the step of task scheduling, separately in each processor. This paper proposes a fuzzy-based technique to distribute real-time tasks among processors in order to reduce both temperature and energy consumption simultaneously. Simulation results show that our proposed technique is more efficient in terms of reducing the energy consumption compared to a well-known state-of-the-art method (up to 9%) while offering a more balanced and moderate temperatures for processors and hindering hot spot. [ABSTRACT FROM AUTHOR]

In the domain of proteomics, an in-depth analysis of the 3D structure of a protein is of paramount importance for many biological studies and applications. At the secondary level, protein structure can be described in terms of motifs, recurrent patterns of smaller biological structures called secondary structure elements. In this paper, the focus is on the identification of geometrical motifs in different proteins using the Cross Motif Search (CMS) algorithm. Such task, due to the high computational cost of CMS with respect to traditional alignment algorithms, is very demanding, and thus, parallel processing is mandatory. In previous papers, CMS parallelization has been already studied from the HPC standpoint. Since cloud computing is emerging as an alternative to on-premise HPC systems, it is worthwhile examining the feasibility and possible advantages in terms of both performance and costs, of migrating to a cloud implementation. This paper is an extension of a preliminary work carried out on the cloud parallelization of CMS. The paper has two main contributions. First of all, an analytic model of the communication pattern of CMS is described, in order to get insights on the performance of the application when executed on a cloud infrastructure. Secondly, an optimized "location-aware" scheduling policy to assign workload to the application workers is introduced, in order to minimize internode communication in a cloud setting. Experiments are presented in order to validate the newly introduced scheduling policy and assess the performance of the cloud implementation of CMS. The results presented in this paper are general, in the sense that they can be applied to any other algorithm with a communication pattern similar to the one of the target applications. [ABSTRACT FROM AUTHOR]

Cloud computing is a popular and widely adopted computing platform for the execution of scientific workflows as it provides flexible infrastructure and offers access to collection of autonomous heterogeneous resources. Effective scheduling of computationally complex workflows which contain many interconnected tasks is a complex problem and becomes more challenging in cloud environment. Optimal solutions can be obtained by considering not only the heterogeneity of computation costs involved, but also by taking into account the communication costs among the tasks in a way that schedule length of the application is reduced. In this paper, we propose a list scheduling heuristic, namely minimal optimistic processing time (MOPT), with optimized duplication approach. The additional feature is introduced for the entry task and is applied only in scenarios in which duplication is more practical and effective. The prioritization phase of the proposed work is based on an optimistic processing time matrix that is used for ranking of the tasks. The algorithm has same time complexity as state-of-the-art existing algorithms, but notable improvements are acquired in terms of makespan and other performance evaluation parameters. Extensive experimental analysis of the proposed algorithm is carried out using synthesized graphs and graphs from the real-world applications. The results prove that MOPT achieves quality schedules with reduced makespans. As communication cost among the tasks grows higher, performance of the proposed algorithm becomes more effective, thus providing the evidence that the MOPT algorithm is well-suited for communication-intensive applications. [ABSTRACT FROM AUTHOR]

Heterogeneous systems are the core architecture of most of the high-performance computing nodes, due to their excellent performance and energy efficiency. However, a key challenge that remains is programmability, specifically, releasing the programmer from the burden of managing data and devices with different architectures. To this end, we extend EngineCL to support FPGA devices. Based on OpenCL, EngineCL is a high-level framework providing load balancing among devices. Our proposal fully integrates FPGAs into the framework, enabling effective cooperation between CPU, GPU, and FPGA. With command overlapping and judicious data management, our work improves performance by up to 96% compared with single-device execution and delivers energy-delay gains of up to 37%. In addition, adopting FPGAs does not require programmers to make big changes in their applications because the extensions do not modify the user-facing interface of EngineCL. [ABSTRACT FROM AUTHOR]

Task-parallel programming models have alleviated the gap between software and hardware complexity in high-performance computing. However, the developer is still in charge of complex decisions that have a significant impact in the overall efficiency and affect the application development. Specifically, in a context in which a set of heterogeneous and interdependent tasks share resources, there is a complex interplay between different factors such as task granularity, task criticality, problem size, application inter-task and intra-task parallelism and available hardware concurrency. In this paper, we explore the effects of this mix from a static scheduling perspective, by exposing a mixed-integer linear program in which the amount of inter- and intra-task parallelism can be adapted as the execution evolves. We solve a set of instances simulating a dense Cholesky factorization on a 20-core Xeon multiprocessor in a power-constrained scenario targeting makespan and energy minimization. The model reveals performance gains up to 17.9% in terms of performance and 4.1% in terms of energy by discovering a set of high-quality scheduling solutions. [ABSTRACT FROM AUTHOR]

Nowadays, cloud computing is a growing scenario applied to many scientific and manufacturing areas due to its flexibility for adapting to highly demanding computing requirements. The advantages of pay-as-you-go model, elasticity, and the flexibility and customization offered by virtualization make cloud computing an attractive option for meeting the needs of some high-performance computing (HPC) users. However, in this environment, the inherent resources heterogeneity, the virtual machine resource sharing, and the HPC-agnostic cloud schedulers are some bottlenecks for effective HPC in cloud. Furthermore, the energy factor has added another layer of complexity in the task scheduling because of the necessity of maximizing the resources utilization and reducing their idle states. In such a complex infrastructure, the scheduling process that allocates the user parallel tasks, represented by cloudlets, to the virtual machines becomes the focus not only to reduce the job execution times, but also to deal with the energy-performance trade-off. In this work, we propose a multi-objective genetic algorithm to determine the most suitable allocation of cloudlets to the available virtual machines. This innovative approach is able to generate scheduling decisions evading systematic allocations and providing new chances for the remaining cloudlets to be scheduled in order to reduce the whole execution time and also the energy consumption. We validated our proposal using real workload traces from HPC environments and compared the results with well-known algorithms from the literature. The obtained results showed that our proposal achieves lower execution times and minimum energy consumption compared with other classic algorithms from the literature. [ABSTRACT FROM AUTHOR]

Cloud schedulers that allocate resources exclusively to single workflows are not work-conserving as they may be forced to leave gaps in their schedules because of the precedence constraints in the workflows. Thus, they may lead to a waste of financial resources. This problem can be mitigated by multiple-workflow schedulers that share the leased cloud resources among multiple workflows or users by filling the gaps left by one workflow with the tasks of other workflows. This solution may even work when users have different performance objectives for their workflows, such as budgets and deadlines. As an additional requirement, we want the scheduler to be fair to all workflows regardless of their performance objectives. In this paper, we propose a multiple-workflow scheduler that is able to target different quality of service goals for different workflows and that considers fairness among different users. To this aim, we propose an unfairness metric and four workflow selection policies. We prove that the resource selection that decides based on a task's sub-budget, sub-deadline, finish time, and cost on different resources is selecting the best resource based on the given information, while using the smallest number of calculations. Simulations show that there is a trade-off between overall cost, makespan, and fairness. We conclude that the best workflow selection policy to reduce unfairness is the direct policy, which explicitly selects the workflow that minimizes the value of the proposed unfairness metric in each round. [ABSTRACT FROM AUTHOR]

Resource provisioning strategies are crucial for workflow scheduling problems which are widespread in cloud computing. The main challenge lies in determining the amounts of reserved and on-demand resources to meet users’ requirements. In this paper, we consider the cloud workflow scheduling problem with hybrid resource provisioning to minimize the total renting cost, which is NP-hard and has not been studied yet. An iterative population-based meta-heuristic is developed. According to the shift vectors obtained during the search procedure, timetables are computed quickly. The appropriate amounts of reserved and on-demand resources are determined by an incremental optimization method. The utilization of each resource is balanced in a swaying way, in terms of which the probabilistic matrix is updated for the next iteration. The proposed algorithm is compared with modified existing algorithms for similar problems. Experimental results demonstrate effectiveness and efficiency of the proposed algorithm. [ABSTRACT FROM AUTHOR]

Time-to-solution is an important metric when parallelizing existing code. The REPARA approach provides a systematic way to instantiate stream and data parallel patterns by annotating the sequential source code with C++11 attributes. Annotations are automatically transformed in a target parallel code that uses existing libraries for parallel programming (e.g., FastFlow). In this paper, we apply this approach for the parallelization of a data stream processing application. The description shows the effectiveness of the approach in easily and quickly prototyping several parallel variants of the sequential code by obtaining good overall performance in terms of both throughput and latency. [ABSTRACT FROM AUTHOR]

The contemporary multicore era has adhered to the heterogeneous computing devices as one of the proficient platforms to execute compute-intensive applications. These heterogeneous devices are based on CPUs and GPUs. OpenCL is deemed as one of the industry standards to program heterogeneous machines. The conventional application scheduling mechanisms allocate most of the applications to GPUs while leaving CPU device underutilized. This underutilization of slower devices (such as CPU) often originates the sub-optimal performance of data-parallel applications in terms of load balance, execution time, and throughput. Moreover, multiple scheduled applications on a heterogeneous system further aggravate the problem of performance inefficiency. This paper is an attempt to evade the aforementioned deficiencies via initiating a novel scheduling strategy named OSched. An enhancement to the OSched named E-OSched is also part of this study. The OSched performs the resource-aware assignment of jobs to both CPUs and GPUs while ensuring a balanced load. The load balancing is achieved via contemplation on computational requirements of jobs and computing potential of a device. The load-balanced execution is beneficiary in terms of lower execution time, higher throughput, and improved utilization. The E-OSched reduces the magnitude of the main memory contention during concurrent job execution phase. The mathematical model of the proposed algorithms is evaluated by comparison of simulation results with different state-of-the-art scheduling heuristics. The results revealed that the proposed E-OSched has performed significantly well than the state-of-the-art scheduling heuristics by obtaining up to 8.09% improved execution time and up to 7.07% better throughput. [ABSTRACT FROM AUTHOR]

Nowadays, we live in a Big Data world and many sectors of our economy are guided by data-driven decision processes. Big Data and Business Intelligence applications are facilitated by the MapReduce programming model, while, at infrastructural layer, cloud computing provides flexible and cost-effective solutions to provide on-demand large clusters. Capacity allocation in such systems, meant as the problem of providing computational power to support concurrent MapReduce applications in a cost-effective fashion, represents a challenge of paramount importance. In this paper we lay the foundation for a solution implementing admission control and capacity allocation for MapReduce jobs with a priori deadline guarantees. In particular, shared Hadoop 2.x clusters supporting batch and/or interactive jobs are targeted. We formulate a linear programming model able to minimize cloud resources costs and rejection penalties for the execution of jobs belonging to multiple classes with deadline guarantees. Scalability analyses demonstrated that the proposed method is able to determine the global optimal solution of the linear problem for systems including up to 10,000 classes in less than 1 s. [ABSTRACT FROM AUTHOR]

In hybrid geo-distributed clouds, there is a technique named cloud bursting in which applications are handled in the private cloud with less expenses and burst into public clouds when the resources of the private cloud run out. However, how to deploy heterogeneous jobs in heterogeneous hybrid cloud environment is still a challenge. In this paper, a multi-queue scheduling approach of heterogeneous jobs for cloud bursting is proposed. In the private cloud, jobs are classified into I/O-intensive and CPU-intensive jobs, and nodes are divided into main I/O and CPU resource pools. Jobs are dispatched to corresponding resource pools to reduce the job execution time in heterogeneous cloud environment. A genetic algorithm is applied to schedule jobs to optimal job queues, which can reduce the job waiting time. Then, the execution time of each task is predicted by BP neural network. Jobs with high priority will be allocated to resources with the earliest finish time in the private cloud according to the prediction results. If the private cloud cannot meet the demand of users, public clouds with minimal costs will be applied. Experiments show that our proposed algorithm can reduce the average job response time and improve the throughput of the private cloud. It also can reduce the average task waiting time, average task execution time and average task response time significantly. Moreover, the costs of the hybrid clouds are reduced. [ABSTRACT FROM AUTHOR]

As cloud computing is a market-oriented utility, optimal virtual machine (VM) scheduling in cloud computing should take into account the incentives for both cloud users and the cloud provider. However, most of existing studies on VM scheduling only consider the incentive for one party, i.e., either the cloud users or the cloud provider. Very few related studies consider the incentives for both parties, in which the cost, one of the most attractive incentives for cloud users, is not well addressed. In this paper, we investigate the problem of VM scheduling in cloud computing by optimizing the incentives for both parties. The problem is formulated as a multi-objective optimization model, i.e., maximizing the successful execution rate of VM requests and minimizing the combined cost (incentives for cloud users), and minimizing the fairness deviation of profits (incentive for the cloud provider). The proposed multi-objective optimization model can offer sufficient incentives for the two parties to stay and play in the cloud and keep the cloud system sustainable. A heuristic-based scheduling algorithm, called cost-greedy dynamic price scheduling, is then developed to optimize the incentives for both parties. Experimental results show that, compared with some popular algorithms, the developed algorithm can achieve higher successful execution rate, lower execution cost, smaller fairness deviation and most important, higher degree of user satisfaction in most cases. [ABSTRACT FROM AUTHOR]

The task scheduling problem on heterogeneous computing systems has been broadly studied, and many heuristic algorithms are proposed to solve this problem. It is interesting to go for significant performance assessment and comparison among these heuristic algorithms. In this work, we first carry out performance evaluation and comparison of task scheduling algorithms for heterogeneous computing systems for randomly generated graphs and the graphs generated from real-world applications such as Fast Fourier Transform, Gaussian Elimination, Montage and Epigenomics workflow. Further, we explores possibility of a framework for benchmarking of task scheduling algorithms for heterogeneous computing systems. This proposed approach provides for generation of graphs through a Directed Acyclic Graph generator, then produces schedules through a scheduler which makes use of scheduling algorithms and finally analyses the results obtained by using various performance metrics. The proposed framework is general in nature. [ABSTRACT FROM AUTHOR]

In recent years, deploying and running data-intensive workflows in cloud platform has become more and more popular in many areas. Unlike computation-intensive applications, a data-intensive workflow typically requires to deal with bulk data transferring between different resource sites, which means some traditional energy-efficiency optimization technologies are difficult to be enforced when running data-intensive workflows. In this paper, we first formulate the power model of a data-intensive workflow, which takes into account power consumption caused by data transferring. Based on this power model, we introduce a novel metric called Shortest Path in terms of Energy Consumption and design an energy-efficient heuristic scheduling algorithm, which is aiming at reducing the extra energy consumption caused by delays of bulk data transferring. Extensive experiments and performance evaluations show that the proposed scheduling algorithm can significantly reduce the overall energy consumption of running data-intensive workflows comparing with several existing algorithms. In addition, the proposed algorithm also exhibits better adaptiveness and robustness when a cloud system is facing intensive and unpredicted workloads. [ABSTRACT FROM AUTHOR]

Although virtualization technology is recently applied to next-generation distributed high-performance computing systems, theoretical aspects of scheduling jobs on these virtualized environments are not sufficiently studied, especially in online and non-clairvoyant cases. Virtualization of computing resources results in interference and virtualization overheads that negatively impact the load balancing objectives on commonly used cluster of multi-core physical machines. We present a technique for non-clairvoyant online scheduling of globally synchronized jobs, each of which spawns tasks to execute compute-intensive works. Our technique considers both load balancing of physical cores and per job synchronization cost minimization. We show that in the presence of arbitrary virtualization overheads, interference effects and synchronization cost, the problem can be reduced to an online unrelated parallel machine scheduling, which is solved using routing of virtual circuits. We present a new opportunity cost model to reduce the problem to the routing of virtual circuits and prove the effectiveness of our scheduling technique using mathematical analysis and simulative experiments. [ABSTRACT FROM AUTHOR]

In this paper, a hybrid scheduling and mapping approach to jointly optimize performance, lifetime reliability, energy consumption and temperature of heterogeneous multiprocessor systems on chip (MPSoCs), called “HYSTERY,” is proposed. Due to the growth of dynamic behavior in modern applications of embedded systems, along with necessity of performing complicated computations to jointly optimize the main design challenges of MPSoCs, we propose a hybrid scheduling approach in this paper. The proposed approach deals with the optimization of the mentioned design challenges at the design-time through solving an optimization problem and considering load balancing in task assignment. Moreover, at the runtime, the derived static solution is applied to the system and the design metrics monitored periodically and controlled, if required, to adapt the static scheduling decisions at the runtime. Several experiments with synthetic and real-life applications demonstrate that the proposed approach can effectively optimize the design challenges and manage dynamism of execution environment. In comparison with the uncontrolled runtime scheduling approach, HYSTERY shows 20% improvement in temperature averagely, which subsequently enhance lifetime reliability and power consumption. Furthermore, HYSTERY improves the main design parameters of MPSOCs about 21% in average compared to the existing scheduling approaches. [ABSTRACT FROM AUTHOR]

Cloud computing allows execution and deployment of different types of applications such as interactive databases or web-based services which require distinctive types of resources. These applications lease cloud resources for a considerably long period and usually occupy various resources to maintain a high quality of service (QoS) factor. On the other hand, general big data batch processing workloads are less QoS-sensitive and require massively parallel cloud resources for short period. Despite the elasticity feature of cloud computing, fine-scale characteristics of cloud-based applications may cause temporal low resource utilization in the cloud computing systems, while process-intensive highly utilized workload suffers from performance issues. Therefore, ability of utilization efficient scheduling of heterogeneous workload is one challenging issue for cloud owners. In this paper, addressing the heterogeneity issue impact on low utilization of cloud computing system, conjunct resource allocation scheme of cloud applications and processing jobs is presented to enhance the cloud utilization. The main idea behind this paper is to apply processing jobs and cloud applications jointly in a preemptive way. However, utilization efficient resource allocation requires exact modeling of workloads. So, first, a novel methodology to model the processing jobs and other cloud applications is proposed. Such jobs are modeled as a collection of parallel and sequential tasks in a Markovian process. This enables us to analyze and calculate the efficient resources required to serve the tasks. The next step makes use of the proposed model to develop a preemptive scheduling algorithm for the processing jobs in order to improve resource utilization and its associated costs in the cloud computing system. Accordingly, a preemption-based resource allocation architecture is proposed to effectively and efficiently utilize the idle reserved resources for the processing jobs in the cloud paradigms. Then, performance metrics such as service time for the processing jobs are investigated. The accuracy of the proposed analytical model and scheduling analysis is verified through simulations and experimental results. The simulation and experimental results also shed light on the achievable QoS level for the preemptively allocated processing jobs. [ABSTRACT FROM AUTHOR]

In recent years, building the cloud based on wireless mesh networks as well as wired networks is rapidly increased for processing the big data. However, existing scheduling and routing protocols cannot support processing the big data efficiently in the cloud, because the each flow path is determined before the data are transmitted based on some routing strategies in the wireless mesh networks. Currently, an important factor that should be considered is that the link capacity between mesh routers can be changed based on the current interference of the flow of other mesh routers. In general, network availability is also influenced by the interference related to the other layers in the protocol stack. In this paper, we study wireless mesh networks and propose JRS-S and JRS-M algorithms,which utilize both route discovery and resource allocation at the same time, in order to maximize capacity of the wireless mesh network in the cloud computing. Our algorithms for each flow use a cross-layer design method based on numerical modeling in order to adaptively control data scheduling at the link layer and find a high data rate path with minimum interference at the network layer.We analyze the optimal capacity of the wireless mesh networks for maximizing network utilization using a numerical solution tool. Through analysis, we also verify that our algorithms can improve system capacity by efficiently distributing a gateway load and that it can enhance the system availability. [ABSTRACT FROM AUTHOR]

Earliest deadline first (EDF) scheduling algorithm is the most celebrated result for dynamic priority scheduling in real-time systems for both preemptive and non-preemptive cases. From complexity point of view, EDF is polynomial for preemptive scheduling of tasks; however, it becomes pseudo-polynomial under non-preemptive case. In this paper, we propose a technique that determines EDF feasibility of non-preemptive task set by analyzing schedulability of the lowest priority task at common scheduling points generated by all higher priority tasks in the task set. This adjustment results in improving the computational cost of an existing test from $$O(n^2 p_n/p_1)$$ to $$O(p_n/p_1)$$ , where n is the number of tasks in the system, while $$p_n$$ and $$p_1$$ represent the task periods of largest and smallest periodic tasks respectively. With reduced computation cost, we understand that our technique has the potential to be intergraded with online systems for testing feasibility of a special class of real-time systems under non-preemptive case. [ABSTRACT FROM AUTHOR]

The MapReduce framework is considered to be an effective resolution for huge and parallel data processing. This paper treats a massive data processing workflow as a DAG graph consisting of MapReduce jobs. In a heterogeneous computing environment, the computation speed can be different even on the same slot depending on various jobs. For this problem, this paper proposes an optimized MapReduce workflow scheduling algorithm. This algorithm comprises a job prioritizing phase and a task assignment phase. First, the jobs can be classified as I/O-intensive and computing-intensive, and the priorities of all jobs are computed according to their corresponding types. Then, the suitable slots are allocated for each block, and the MapReduce tasks in the workflow are scheduled with respect to data locality. The experimental results show that the optimized MapReduce workflow scheduling algorithm can improve the performance of task scheduling and the rationality of resources allocation in heterogeneous computing. [ABSTRACT FROM AUTHOR]

Cloud computing provides users a shared pool of configurable computing resources. In this paper, a cloud computing system is regarded as a queuing system, where users arrive according to a stochastic process and request resources, including CPU, memory, storage space. etc. To improve the utilization of the system under stable state, we provide some theoretical results about the relationship between the utilization and the stability of the cloud computing system. The conditions for the system to be stable are given for systems with preemptive priority and non-preemptive priority, respectively. Given the stability conditions, we suggest a scheduling algorithm to improve the optimal utilization of the could computing system with preemptive priority and non-preemptive priority, respectively. Numerical results indicate that the two algorithms provide adequate performance on utilization. In addition, the influence of different parameters on the algorithm is investigated as well. [ABSTRACT FROM AUTHOR]

Large-scale MapReduce clusters that routinely process big data bring challenges to the cloud computing. One of the key challenges is to reduce the response time of these MapReduce clusters by minimizing their makespans. It is observed that the order in which these jobs are executed can have a significant impact on their overall makespans and resource utilization. In this work, we consider a scheduling model for multiple MapReduce jobs. The goal is to design a job scheduler that minimizes the makespan of such a set of MapReduce jobs. We exploit classical Johnson model and propose a novel framework HScheduler, which combines features of both classical Johnson's algorithm and MapReduce to minimize the makespan for both offline and online jobs. Our Offline HScheduler reaches the theoretical lower bound (optimum) and Online HScheduler is 2-competitive which is the best-known constant ratio for minimizing the makespan. Through extensive real data tests, we find that HScheduler has better performance than the best-known approach by 10.6-11.7 % on average for offline scheduling and 8-10 % on average for online scheduling. The HScheduler can be applied to improve responsive time, throughput and energy efficiency in cloud computing. [ABSTRACT FROM AUTHOR]

Current compilers cannot generate code that can compete with hand-tuned code in efficiency, even for a simple kernel like matrix-matrix multiplication (MMM). A key step in program optimization is the estimation of optimal values for parameters such as tile sizes and number of levels of tiling. The scheduling parameter values selection is a very difficult and time-consuming task, since parameter values depend on each other; this is why they are found by using searching methods and empirical techniques. To overcome this problem, the scheduling sub-problems must be optimized together, as one problem and not separately. In this paper, an MMM methodology is presented where the optimum scheduling parameters are found by decreasing the search space theoretically, while the major scheduling sub-problems are addressed together as one problem and not separately according to the hardware architecture parameters and input size; for different hardware architecture parameters and/or input sizes, a different implementation is produced. This is achieved by fully exploiting the software characteristics (e.g., data reuse) and hardware architecture parameters (e.g., data caches sizes and associativities), giving high-quality solutions and a smaller search space. This methodology refers to a wide range of CPU and GPU architectures. [ABSTRACT FROM AUTHOR]

Directed acyclic graph (DAG) scheduling is a well-known problem, because a DAG can be used to describe a wide range of complex applications, including scientific applications and parallel computing jobs. Most DAG scheduling algorithms were proposed to minimize the job makespan (i.e., execution time) on a multiprocessor computer or cluster. However, as the cost-driven public cloud services have become an attractive and popular platform for providing computing resources, cost minimization emerges as a new critical issue. Therefore, the objective of this work is to propose and solve the cost optimization problem for scheduling DAGs on an IaaS cloud platform where task scheduling must cope with resource provisioning to achieve the optimal solution. In this paper, we proposed both optimal and heuristic scheduling algorithms, and we evaluated them across a variety of DAGs using the price model from EC2. Comparing to other cost-oblivious DAG schedules that aim to minimize makespan or resource usage, the results show that our cost-aware heuristic algorithm can reduce cost by 20-50 % and achieve a cost within x1.16 of the optimal one. [ABSTRACT FROM AUTHOR]

Resource provisioning of appropriate resources to cloud workloads depends on the quality of service (QoS) requirements of cloud applications and is a challenging task. In cloud environment, heterogeneity, uncertainty and dispersion of resources encounter a problem of allocation of resources, which cannot be addressed with existing resource management frameworks. Resource scheduling, if done after efficient resource provisioning, will be more effective and the cloud resources would be scheduled as per the user requirements (QoS) on provisioned resources. Execution of cloud workloads should be as per QoS parameters to fully satisfy the cloud consumer. Therefore, based on QoS parameters, it is mandatory to predict and verify the resource provisioning before actual resource scheduling. In this paper, a resource provisioning and scheduling framework has been presented which caters to provisioned resource distribution and scheduling of resources. Cloud workloads have been re-clustered using k-means-based clustering algorithm after firstly clustering them through workload patterns to identify the QoS requirements of a workload, and then based on identified QoS requirements resources are provisioned before actual scheduling. Further, scheduling has been done based on different scheduling policies. Finally, the performance of the proposed framework has been evaluated in both real and simulated cloud environment and experimental results show that the framework provisions and schedules resource efficiently by considering energy consumption, execution cost and execution time as QoS parameters. [ABSTRACT FROM AUTHOR]

The accurate and quantitative analysis of the cache behavior in a Chip Multi-Core (CMP) machine has long been a challenging work. So far there has been no practical way to predict the cache allocation, i.e., allocated cache size, of a running program. Lots of applications, especially those that have many interactions with the users, cache allocation should be estimated with high accuracy since its variation is closely related to the stability of system performance which is important to the efficient operation of servers and has a great influence on user experience. For these interests, this paper proposes an accurate prediction model for the allocation of the last level cache (LLC) of the co-runners. With a precise cache allocation predicted, we further implemented a performance-stability-oriented co-runner scheduling algorithm which aims to maximize the number of co-runners running in performance-stable state and minimize the performance variation of the unstable ones. We demonstrate that the proposed prediction algorithm exhibits a high accuracy with an average error of 5.7 %; and the co-runner scheduling algorithm can find the optimal solution under the specified target with a time complexity of O( n). [ABSTRACT FROM AUTHOR]

The All-optical Linear Array with a Reconfigurable Pipelined Bus System (OLARPBS) optical bus parallel computing model is proposed in this paper. The OLARPBS model includes several architectural and logical extensions to the existing LARPBS(p) model. Architecturally, the extensions include the replacement of the electronic processors by fine-grained, all-optical digital processing elements; and the replacement of the optical conduit (optical bus) by one or more optical conduits, each consisting of a structural hierarchy of bundles of light paths that enable multiple parallel and/or pipelined message communications. Logically, the extensions include: parallel and/or pipelined, conduit-based, optical registers; register scheduling; and processor data path requirements. Collectively, these extensions enable register-based bits-in-flight algorithm development. A study of two applications with implementation on the OLARPBS model is considered. [ABSTRACT FROM AUTHOR]

Software transactional memory (STM) is one of the techniques used towards achieving non-blocking process synchronization in multi-threaded computing environment. In spite of its high potential, one of the major limitations of transactional memory (TM) is that in order to ensure data consistency as well as progress condition, TM often forces transactions to abort. This paper proposes a new concurrency control mechanism. It starts with the existing TM implementations for obstruction freedom and eventually builds a new STM methodology. The primary objective is to reduce aborting of transactions in some typical scenarios. A programming model is described for a chain of update transactions that share the same data object among themselves. Using the proposed approach, any new update transaction appended in this chain need not wait for the earlier transactions to finish. The proposed STM allows wait-free, non-blocking implementation of a mix of read and multiple update transactions on the same shared data object with higher throughput. [ABSTRACT FROM AUTHOR]